| Message ID | 1663747240-31210-1-git-send-email-wangyufen@huawei.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | BPF |
| Headers | show |
| Series | [bpf-next,v5,1/3] bpftool: Add auto_attach for bpf prog load|loadall | expand |
| Context | Check | Description |
|---|---|---|
| netdev/tree_selection | success | Clearly marked for bpf-next |
| netdev/fixes_present | success | Fixes tag not required for -next series |
| netdev/subject_prefix | success | Link |
| netdev/cover_letter | warning | Series does not have a cover letter |
| netdev/patch_count | success | Link |
| netdev/header_inline | success | No static functions without inline keyword in header files |
| netdev/build_32bit | success | Errors and warnings before: 0 this patch: 0 |
| netdev/cc_maintainers | success | CCed 17 of 17 maintainers |
| netdev/build_clang | success | Errors and warnings before: 0 this patch: 0 |
| netdev/module_param | success | Was 0 now: 0 |
| netdev/verify_signedoff | success | Signed-off-by tag matches author and committer |
| netdev/check_selftest | success | No net selftest shell script |
| netdev/verify_fixes | success | No Fixes tag |
| netdev/build_allmodconfig_warn | success | Errors and warnings before: 0 this patch: 0 |
| netdev/checkpatch | success | total: 0 errors, 0 warnings, 0 checks, 113 lines checked |
| netdev/kdoc | success | Errors and warnings before: 0 this patch: 0 |
| netdev/source_inline | success | Was 0 now: 0 |
| bpf/vmtest-bpf-next-VM_Test-1 | success | Logs for build for s390x with gcc |
| bpf/vmtest-bpf-next-VM_Test-16 | success | Logs for test_verifier on x86_64 with gcc |
| bpf/vmtest-bpf-next-VM_Test-17 | success | Logs for test_verifier on x86_64 with llvm-16 |
| bpf/vmtest-bpf-next-VM_Test-10 | success | Logs for test_progs on x86_64 with gcc |
| bpf/vmtest-bpf-next-VM_Test-13 | success | Logs for test_progs_no_alu32 on x86_64 with gcc |
| bpf/vmtest-bpf-next-VM_Test-14 | success | Logs for test_progs_no_alu32 on x86_64 with llvm-16 |
| bpf/vmtest-bpf-next-VM_Test-11 | success | Logs for test_progs on x86_64 with llvm-16 |
| bpf/vmtest-bpf-next-VM_Test-7 | success | Logs for test_maps on x86_64 with gcc |
| bpf/vmtest-bpf-next-VM_Test-8 | success | Logs for test_maps on x86_64 with llvm-16 |
| bpf/vmtest-bpf-next-VM_Test-6 | fail | Logs for test_maps on s390x with gcc |
| bpf/vmtest-bpf-next-VM_Test-15 | success | Logs for test_verifier on s390x with gcc |
| bpf/vmtest-bpf-next-VM_Test-12 | success | Logs for test_progs_no_alu32 on s390x with gcc |
| bpf/vmtest-bpf-next-PR | fail | PR summary |
| bpf/vmtest-bpf-next-VM_Test-9 | success | Logs for test_progs on s390x with gcc |
| bpf/vmtest-bpf-next-VM_Test-2 | success | Logs for build for x86_64 with gcc |
| bpf/vmtest-bpf-next-VM_Test-3 | success | Logs for build for x86_64 with llvm-16 |
| bpf/vmtest-bpf-next-VM_Test-4 | success | Logs for llvm-toolchain |
| bpf/vmtest-bpf-next-VM_Test-5 | success | Logs for set-matrix |
On Wed, Sep 21, 2022 at 12:40 AM Wang Yufen <wangyufen@huawei.com> wrote: > > Add auto_attach optional to support one-step load-attach-pin_link. > > For example, > $ bpftool prog loadall test.o /sys/fs/bpf/test auto_attach > > $ bpftool link > 26: tracing name test1 tag f0da7d0058c00236 gpl > loaded_at 2022-09-09T21:39:49+0800 uid 0 > xlated 88B jited 55B memlock 4096B map_ids 3 > btf_id 55 > 28: kprobe name test3 tag 002ef1bef0723833 gpl > loaded_at 2022-09-09T21:39:49+0800 uid 0 > xlated 88B jited 56B memlock 4096B map_ids 3 > btf_id 55 > 57: tracepoint name oncpu tag 7aa55dfbdcb78941 gpl > loaded_at 2022-09-09T21:41:32+0800 uid 0 > xlated 456B jited 265B memlock 4096B map_ids 17,13,14,15 > btf_id 82 > > $ bpftool link > 1: tracing prog 26 > prog_type tracing attach_type trace_fentry > 3: perf_event prog 28 > 10: perf_event prog 57 > > The auto_attach optional can support tracepoints, k(ret)probes, > u(ret)probes. > > Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com> > Signed-off-by: Wang Yufen <wangyufen@huawei.com> > Reviewed-by: Quentin Monnet <quentin@isovalent.com> > --- > v4 -> v5: some formatting nits of doc > v3 -> v4: rename functions, update doc, bash and do_help() > v2 -> v3: switch to extend prog load command instead of extend perf > v2: https://patchwork.kernel.org/project/netdevbpf/patch/20220824033837.458197-1-weiyongjun1@huawei.com/ > v1: https://patchwork.kernel.org/project/netdevbpf/patch/20220816151725.153343-1-weiyongjun1@huawei.com/ > tools/bpf/bpftool/prog.c | 77 ++++++++++++++++++++++++++++++++++++++++++++++-- > 1 file changed, 75 insertions(+), 2 deletions(-) > > diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c > index c81362a..aea0b57 100644 > --- a/tools/bpf/bpftool/prog.c > +++ b/tools/bpf/bpftool/prog.c > @@ -1453,6 +1453,68 @@ static int do_run(int argc, char **argv) > return ret; > } > > +static int > +auto_attach_program(struct bpf_program *prog, const char *path) > +{ > + struct bpf_link *link; > + int err; > + > + link = bpf_program__attach(prog); > + err = libbpf_get_error(link); nit: bpftool uses libbpf 1.0, so no need to use libbpf_get_error() anymore, you can just check link for NULL and then look at errno but I wanted to check on desired behavior here. BPF skeleton will skip programs that can't be auto-attached because they are of the type that can't be declaratively specified to be auto-attachable. For such programs bpf_program__attach() will return -EOPNOTSUPP and libbpf's skeleton_attach API will silently skip them. Should bpftool be stricter about such programs here or should it follow BPF skeleton approach? > + if (err) > + return err; > + > + err = bpf_link__pin(link, path); > + if (err) { > + bpf_link__destroy(link); > + return err; > + } > + return 0; > +} > + > +static int pathname_concat(const char *path, const char *name, char *buf) you added buffer size in libbpf version of this function, maybe match the same signature (I also moved buf and buf_sz to be first two args). > +{ > + int len; > + > + len = snprintf(buf, PATH_MAX, "%s/%s", path, name); > + if (len < 0) > + return -EINVAL; > + if (len >= PATH_MAX) > + return -ENAMETOOLONG; > + > + return 0; > +} > + > +static int > +auto_attach_programs(struct bpf_object *obj, const char *path) > +{ > + struct bpf_program *prog; > + char buf[PATH_MAX]; > + int err; > + > + bpf_object__for_each_program(prog, obj) { > + err = pathname_concat(path, bpf_program__name(prog), buf); > + if (err) > + goto err_unpin_programs; > + > + err = auto_attach_program(prog, buf); > + if (err) > + goto err_unpin_programs; > + } > + would it be safer to first make sure that all programs are auto-attached and then pin links? also note that not all bpf_links returned by libbpf are actual links in kernel (e.g., kprobe/tp bpf_link on older kernels). > + return 0; > + > +err_unpin_programs: > + while ((prog = bpf_object__prev_program(obj, prog))) { > + if (pathname_concat(path, bpf_program__name(prog), buf)) > + continue; > + > + bpf_program__unpin(prog, buf); > + } > + > + return err; > +} > + > static int load_with_options(int argc, char **argv, bool first_prog_only) > { > enum bpf_prog_type common_prog_type = BPF_PROG_TYPE_UNSPEC; > @@ -1464,6 +1526,7 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) > struct bpf_program *prog = NULL, *pos; > unsigned int old_map_fds = 0; > const char *pinmaps = NULL; > + bool auto_attach = false; > struct bpf_object *obj; > struct bpf_map *map; > const char *pinfile; > @@ -1583,6 +1646,9 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) > goto err_free_reuse_maps; > > pinmaps = GET_ARG(); > + } else if (is_prefix(*argv, "auto_attach")) { > + auto_attach = true; > + NEXT_ARG(); > } else { > p_err("expected no more arguments, 'type', 'map' or 'dev', got: '%s'?", > *argv); > @@ -1692,14 +1758,20 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) > goto err_close_obj; > } > > - err = bpf_obj_pin(bpf_program__fd(prog), pinfile); > + if (auto_attach) > + err = auto_attach_program(prog, pinfile); > + else > + err = bpf_obj_pin(bpf_program__fd(prog), pinfile); > if (err) { > p_err("failed to pin program %s", > bpf_program__section_name(prog)); > goto err_close_obj; > } > } else { > - err = bpf_object__pin_programs(obj, pinfile); > + if (auto_attach) > + err = auto_attach_programs(obj, pinfile); > + else > + err = bpf_object__pin_programs(obj, pinfile); > if (err) { > p_err("failed to pin all programs"); > goto err_close_obj; > @@ -2338,6 +2410,7 @@ static int do_help(int argc, char **argv) > " [type TYPE] [dev NAME] \\\n" > " [map { idx IDX | name NAME } MAP]\\\n" > " [pinmaps MAP_DIR]\n" > + " [auto_attach]\n" looking at "pinmaps" seems like "autoattach" would be more consistent naming? Or just "attach"? > " %1$s %2$s attach PROG ATTACH_TYPE [MAP]\n" > " %1$s %2$s detach PROG ATTACH_TYPE [MAP]\n" > " %1$s %2$s run PROG \\\n" > -- > 1.8.3.1 >
在 2022/9/24 5:29, Andrii Nakryiko 写道: > On Wed, Sep 21, 2022 at 12:40 AM Wang Yufen <wangyufen@huawei.com> wrote: >> Add auto_attach optional to support one-step load-attach-pin_link. >> >> For example, >> $ bpftool prog loadall test.o /sys/fs/bpf/test auto_attach >> >> $ bpftool link >> 26: tracing name test1 tag f0da7d0058c00236 gpl >> loaded_at 2022-09-09T21:39:49+0800 uid 0 >> xlated 88B jited 55B memlock 4096B map_ids 3 >> btf_id 55 >> 28: kprobe name test3 tag 002ef1bef0723833 gpl >> loaded_at 2022-09-09T21:39:49+0800 uid 0 >> xlated 88B jited 56B memlock 4096B map_ids 3 >> btf_id 55 >> 57: tracepoint name oncpu tag 7aa55dfbdcb78941 gpl >> loaded_at 2022-09-09T21:41:32+0800 uid 0 >> xlated 456B jited 265B memlock 4096B map_ids 17,13,14,15 >> btf_id 82 >> >> $ bpftool link >> 1: tracing prog 26 >> prog_type tracing attach_type trace_fentry >> 3: perf_event prog 28 >> 10: perf_event prog 57 >> >> The auto_attach optional can support tracepoints, k(ret)probes, >> u(ret)probes. >> >> Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com> >> Signed-off-by: Wang Yufen <wangyufen@huawei.com> >> Reviewed-by: Quentin Monnet <quentin@isovalent.com> >> --- >> v4 -> v5: some formatting nits of doc >> v3 -> v4: rename functions, update doc, bash and do_help() >> v2 -> v3: switch to extend prog load command instead of extend perf >> v2: https://patchwork.kernel.org/project/netdevbpf/patch/20220824033837.458197-1-weiyongjun1@huawei.com/ >> v1: https://patchwork.kernel.org/project/netdevbpf/patch/20220816151725.153343-1-weiyongjun1@huawei.com/ >> tools/bpf/bpftool/prog.c | 77 ++++++++++++++++++++++++++++++++++++++++++++++-- >> 1 file changed, 75 insertions(+), 2 deletions(-) >> >> diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c >> index c81362a..aea0b57 100644 >> --- a/tools/bpf/bpftool/prog.c >> +++ b/tools/bpf/bpftool/prog.c >> @@ -1453,6 +1453,68 @@ static int do_run(int argc, char **argv) >> return ret; >> } >> >> +static int >> +auto_attach_program(struct bpf_program *prog, const char *path) >> +{ >> + struct bpf_link *link; >> + int err; >> + >> + link = bpf_program__attach(prog); >> + err = libbpf_get_error(link); > nit: bpftool uses libbpf 1.0, so no need to use libbpf_get_error() > anymore, you can just check link for NULL and then look at errno Thanks, will change in v6 > > but I wanted to check on desired behavior here. BPF skeleton will skip > programs that can't be auto-attached because they are of the type that > can't be declaratively specified to be auto-attachable. For such > programs bpf_program__attach() will return -EOPNOTSUPP and libbpf's > skeleton_attach API will silently skip them. Should bpftool be > stricter about such programs here or should it follow BPF skeleton > approach? will change auto_attach_programs() to follow BPF skeleton approach in v6 >> + if (err) >> + return err; >> + >> + err = bpf_link__pin(link, path); >> + if (err) { >> + bpf_link__destroy(link); >> + return err; >> + } >> + return 0; >> +} >> + >> +static int pathname_concat(const char *path, const char *name, char *buf) > you added buffer size in libbpf version of this function, maybe match > the same signature (I also moved buf and buf_sz to be first two args). > >> +{ >> + int len; >> + >> + len = snprintf(buf, PATH_MAX, "%s/%s", path, name); >> + if (len < 0) >> + return -EINVAL; >> + if (len >= PATH_MAX) >> + return -ENAMETOOLONG; >> + >> + return 0; >> +} >> + >> +static int >> +auto_attach_programs(struct bpf_object *obj, const char *path) >> +{ >> + struct bpf_program *prog; >> + char buf[PATH_MAX]; >> + int err; >> + >> + bpf_object__for_each_program(prog, obj) { >> + err = pathname_concat(path, bpf_program__name(prog), buf); >> + if (err) >> + goto err_unpin_programs; >> + >> + err = auto_attach_program(prog, buf); >> + if (err) >> + goto err_unpin_programs; >> + } >> + > would it be safer to first make sure that all programs are > auto-attached and then pin links? > > also note that not all bpf_links returned by libbpf are actual links > in kernel (e.g., kprobe/tp bpf_link on older kernels). will silently skip the unsupport programs as BPF skeleton approach > >> + return 0; >> + >> +err_unpin_programs: >> + while ((prog = bpf_object__prev_program(obj, prog))) { >> + if (pathname_concat(path, bpf_program__name(prog), buf)) >> + continue; >> + >> + bpf_program__unpin(prog, buf); >> + } >> + >> + return err; >> +} >> + >> static int load_with_options(int argc, char **argv, bool first_prog_only) >> { >> enum bpf_prog_type common_prog_type = BPF_PROG_TYPE_UNSPEC; >> @@ -1464,6 +1526,7 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) >> struct bpf_program *prog = NULL, *pos; >> unsigned int old_map_fds = 0; >> const char *pinmaps = NULL; >> + bool auto_attach = false; >> struct bpf_object *obj; >> struct bpf_map *map; >> const char *pinfile; >> @@ -1583,6 +1646,9 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) >> goto err_free_reuse_maps; >> >> pinmaps = GET_ARG(); >> + } else if (is_prefix(*argv, "auto_attach")) { >> + auto_attach = true; >> + NEXT_ARG(); >> } else { >> p_err("expected no more arguments, 'type', 'map' or 'dev', got: '%s'?", >> *argv); >> @@ -1692,14 +1758,20 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) >> goto err_close_obj; >> } >> >> - err = bpf_obj_pin(bpf_program__fd(prog), pinfile); >> + if (auto_attach) >> + err = auto_attach_program(prog, pinfile); >> + else >> + err = bpf_obj_pin(bpf_program__fd(prog), pinfile); >> if (err) { >> p_err("failed to pin program %s", >> bpf_program__section_name(prog)); >> goto err_close_obj; >> } >> } else { >> - err = bpf_object__pin_programs(obj, pinfile); >> + if (auto_attach) >> + err = auto_attach_programs(obj, pinfile); >> + else >> + err = bpf_object__pin_programs(obj, pinfile); >> if (err) { >> p_err("failed to pin all programs"); >> goto err_close_obj; >> @@ -2338,6 +2410,7 @@ static int do_help(int argc, char **argv) >> " [type TYPE] [dev NAME] \\\n" >> " [map { idx IDX | name NAME } MAP]\\\n" >> " [pinmaps MAP_DIR]\n" >> + " [auto_attach]\n" > looking at "pinmaps" seems like "autoattach" would be more consistent > naming? Or just "attach"? will change to "autoattach" in v6 > >> " %1$s %2$s attach PROG ATTACH_TYPE [MAP]\n" >> " %1$s %2$s detach PROG ATTACH_TYPE [MAP]\n" >> " %1$s %2$s run PROG \\\n" >> -- >> 1.8.3.1 >>
diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index c81362a..aea0b57 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -1453,6 +1453,68 @@ static int do_run(int argc, char **argv) return ret; } +static int +auto_attach_program(struct bpf_program *prog, const char *path) +{ + struct bpf_link *link; + int err; + + link = bpf_program__attach(prog); + err = libbpf_get_error(link); + if (err) + return err; + + err = bpf_link__pin(link, path); + if (err) { + bpf_link__destroy(link); + return err; + } + return 0; +} + +static int pathname_concat(const char *path, const char *name, char *buf) +{ + int len; + + len = snprintf(buf, PATH_MAX, "%s/%s", path, name); + if (len < 0) + return -EINVAL; + if (len >= PATH_MAX) + return -ENAMETOOLONG; + + return 0; +} + +static int +auto_attach_programs(struct bpf_object *obj, const char *path) +{ + struct bpf_program *prog; + char buf[PATH_MAX]; + int err; + + bpf_object__for_each_program(prog, obj) { + err = pathname_concat(path, bpf_program__name(prog), buf); + if (err) + goto err_unpin_programs; + + err = auto_attach_program(prog, buf); + if (err) + goto err_unpin_programs; + } + + return 0; + +err_unpin_programs: + while ((prog = bpf_object__prev_program(obj, prog))) { + if (pathname_concat(path, bpf_program__name(prog), buf)) + continue; + + bpf_program__unpin(prog, buf); + } + + return err; +} + static int load_with_options(int argc, char **argv, bool first_prog_only) { enum bpf_prog_type common_prog_type = BPF_PROG_TYPE_UNSPEC; @@ -1464,6 +1526,7 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) struct bpf_program *prog = NULL, *pos; unsigned int old_map_fds = 0; const char *pinmaps = NULL; + bool auto_attach = false; struct bpf_object *obj; struct bpf_map *map; const char *pinfile; @@ -1583,6 +1646,9 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) goto err_free_reuse_maps; pinmaps = GET_ARG(); + } else if (is_prefix(*argv, "auto_attach")) { + auto_attach = true; + NEXT_ARG(); } else { p_err("expected no more arguments, 'type', 'map' or 'dev', got: '%s'?", *argv); @@ -1692,14 +1758,20 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) goto err_close_obj; } - err = bpf_obj_pin(bpf_program__fd(prog), pinfile); + if (auto_attach) + err = auto_attach_program(prog, pinfile); + else + err = bpf_obj_pin(bpf_program__fd(prog), pinfile); if (err) { p_err("failed to pin program %s", bpf_program__section_name(prog)); goto err_close_obj; } } else { - err = bpf_object__pin_programs(obj, pinfile); + if (auto_attach) + err = auto_attach_programs(obj, pinfile); + else + err = bpf_object__pin_programs(obj, pinfile); if (err) { p_err("failed to pin all programs"); goto err_close_obj; @@ -2338,6 +2410,7 @@ static int do_help(int argc, char **argv) " [type TYPE] [dev NAME] \\\n" " [map { idx IDX | name NAME } MAP]\\\n" " [pinmaps MAP_DIR]\n" + " [auto_attach]\n" " %1$s %2$s attach PROG ATTACH_TYPE [MAP]\n" " %1$s %2$s detach PROG ATTACH_TYPE [MAP]\n" " %1$s %2$s run PROG \\\n"